The mechanical left ventricular assist device (LVAD) is often life-saving for patients with severe heart failure. Although LVAD is typically used as a bridge to transplantation, an important subset of patients on LVAD have recovery of LV function leading to successful removal of the device. Assessment of LV function by routine clinical means is difficult or impossible on LVAD due to complex unloading effects. Accordingly, we have developed novel load-independent measures of LV function (preload adjusted maximal power) using on-line digital echocardiography and non-invasive arterial pressure. We have applied these methods to 53 selected LVAD patients, and identified LV recovery in 14 who had successful LVAD removal. This proposal uses a multi-center consortium of 3 expert LVAD centers: the University of Pittsburgh, which will lead all aspects, Penn State-Hershey, and the University of Maryland. The University of Pittsburgh team is a collaboration of cardiac imaging, heart failure and the clinical Artificial Heart Program along with bioengineering experts at Carnegie Mellon University. Aim # 1 will test the hypothesis that our novel loadindependent LV functional measures can reliably identify LV recovery in a series of 175-200 LVAD patients. The predictive power of routine clinical peak exercise oxygen consumption and right heart catheterization will be compared to refine selection criteria for LVAD removal. Specific Aims # 2 and #3 will test the hypothesis that LV recovery on LVAD can be facilitated: Aim #2 will use in vitro and in vivo animal experiments to refine and test a feedback controller for Synchronized LVAD operation for optimal loading to facilitate LV recovery. Aim #3 will test this strategy in a prospective multi-center randomized clinical trial where LVAD Synchronized Adaptive Reloading will be compared to routine Asynchronous LVAD operation. Lav Summary: Mechanical support with a left ventricular assist device (LVAD) can be life-saving for patients with severe heart failure, usually as a bridge to transplantation. Although patients may have cardiac recovery while on LVAD and the device can be removed, it is difficult to assess using routine means. We will use novel non-invasive cardiac testing to identify recovery of the heart on LVAD and develop a unique control system to facilitate cardiac recovery. This Synchronized Adaptive Reloading strategy will be developed using computer models and animal experiments, and will then be tested in a randomized multicenter clinical trial to improve both the efficiency and frequency of cardiac recovery on LVAD.